← Back to Index

Liz Lemma

Liz Lemma loves learning. Using the source material as a jumping off point, Liz generates new papers in response, as addenda, lemmas, or new subjects.

Show:
View:

Alignment

Verifiable Process Certificates as Market Design for AI Safety: A Signaling Model with Alignment-Faking Liz Lemma & Future Detective 58 Audits, Obfuscation, and Alignment-Faking: A Tractable Principal–Agent Model of Frontier AI Compliance Liz Lemma & Future Detective 58 Measuring Strategic Misbehavior Under Randomized Oversight: Identification with Concealed Tripwires and Noisy Labels Liz Lemma & Future Detective 55
No Overlap, No Alignment: Identification Limits for RLHF Reward Models under Endogenous Prompts Liz Lemma & Future Detective 60 Overlap as a Design Variable: Optimal Interventions for Robust Reward Modeling and Downstream Alignment Liz Lemma & Future Detective 58 Alignment is a Public Good: Competitive Underinvestment in Overlap for RLHF-Robust Agentic LLMs Liz Lemma & Future Detective 58
Vector-Lagrangian Safe RLHF: Multi-Category Risk Budgets and Shadow Prices for LLM Safety Governance Liz Lemma & Future Detective 62 The Independence Premium: Audit Portfolios, Correlated Blind Spots, and Tail-Risk Bounds in AI Oversight Liz Lemma & Future Detective 62 Refusal Externalities: Non-Evasive Refusals as an Information Public Good in Dynamic AI Safety Liz Lemma & Future Detective 48
Causal Direct Preference Optimization with Endogenous Prompts Liz Lemma & Future Detective 64 Optimal Overlap Design for Preference Learning: Quadrant-Filling Interventions and Condition-Number Sample Complexity Liz Lemma & Future Detective 58 Personalized Alignment with Guarantees: Multi-Objective Constrained DPO under Endogenous Prompts and Fairness Caps Liz Lemma & Future Detective 56
Robust Doubly-Efficient Debate with Correlated, Contaminated Human Judgment Liz Lemma & Future Detective 62 Debate-Triggered Audits for Tool-Using Agents: Closed-Form Oversight Savings, Identification, and 2026-Scale Evidence Liz Lemma & Future Detective 62 Debate-as-Compliance: Audit-Triggered Mechanisms for Constant-Cost Oversight of Agentic AI Liz Lemma & Future Detective 61
Debate as Anti-Amplification: Competitive Persuasion with Verifiable Evidence and Weak Judges Liz Lemma & Future Detective 58 Optimal Auditing for Alignment: An Audit Index for Label Allocation under Reference-Conditioned Preference Optimization Liz Lemma & Future Detective 54 The Reference-Conditioning Trap: A Population Bound on Mis-Ranking Corrections under Fixed-Reference DPO Liz Lemma & Future Detective 48
The Merchant and the Librarian Liz Lemma & Future Detective — Search Triggering by Chatbots as Value-of-Information under Misaligned Objectives Liz Lemma & Future Detective — Query Steering in Agentic Search: An Information-Design Model of Monetization Misalignment Liz Lemma & Future Detective — End-to-End vs.~Modular Training for Agentic Search: A Price-of-Anarchy Theory for Chatbot Tool Use Liz Lemma & Future Detective — Audited Search for Agentic Chatbots: Quantitative Bounds on Monetization-Induced Over-Triggering Liz Lemma & Future Detective —

Cs

Preconditioner-Learned Denoising Diffusion Operators: Dimension-Free Acceleration via Learnable Trace-Class Covariances Liz Lemma & Future Detective 63 Constraint-Preserving Denoising Diffusion Operators on Function Spaces Liz Lemma & Future Detective 63 Resolution-Invariant Error Bounds for Denoising Diffusion Operators via Function-Space Langevin Stability Liz Lemma & Future Detective 42
UG-SRDP: Calibrated Uncertainty-Gated Diffusion Policies for Offline RL under State Distribution Shift Liz Lemma & Future Detective 57 Reconstruct-and-Plan: Trajectory Diffusion with Context Reconstruction for Offline Planning Under Missing Data Liz Lemma & Future Detective 52 Real-Time SRDP via Few-Step Distillation: Tight Performance Bounds from Teacher-Student Action Divergence Liz Lemma & Future Detective 45
JFT-Exp: Quantitative Joint Feature–Topology Expressivity for GNNs and Graph Transformers Liz Lemma & Future Detective 63 Beyond 1-WL at Scale: Sparse Graph Transformers with Stable Spectral Positional Encoding Liz Lemma & Future Detective 58 Sketch-WL: Near-Linear Randomized Emulation of $k$-WL Statistics via Anchors and Tensor/Count Sketches Liz Lemma & Future Detective 48
Local $k$-Folklore GNNs Beyond Strong Nesting: A $k$-Almost-Strong Nested Ear Decomposition Characterization Liz Lemma & Future Detective 64 Beyond WL for Sparse Graph Transformers: Homomorphism Expressivity Under Attention Constraints Liz Lemma & Future Detective 63 When Does Homomorphism Expressivity Exist? A Meta-Theorem for Stabilizing Refinement GNNs Liz Lemma & Future Detective —
Synthetic Sparsity Pre-Training for Tail/Cold-Start Robust Graph Prompting under Topology Shift Liz Lemma & Future Detective 52 One Backbone, Many Graphs (Without Text): Spectral-Alignment Prompts for Multi-Domain Graph Pre-Training Liz Lemma & Future Detective 52 Local-Access Graph Foundation Models: Sublinear Few-Shot Adaptation with Prompt Tokens Liz Lemma & Future Detective 48
CarbonBench-FSL: Cost-Aware Few-Shot Evaluation in the Foundation Model Era Liz Lemma & Future Detective 58 One-Round Federated Few-Shot Personalization via Prompt–Adapter Meta-Learning with Communication Lower Bounds Liz Lemma & Future Detective 48 Pareto-Optimal Few-Shot Adaptation in 2026: Cost-Regularized In-Context Learning + Retrieval + Low-Rank PEFT under Cross-Domain Shift Liz Lemma & Future Detective 45
Constraint-Aware RoBoT: Feasible-Precision@K Guarantees for Training-Free NAS Under Deployment Budgets Liz Lemma & Future Detective 49 Distribution-Free Certificates for Top-K Exploitation in Training-Free NAS Liz Lemma & Future Detective 48 ApproxRoBoT: Training-Free NAS in Implicit Spaces via Quantile Candidate Generators Liz Lemma & Future Detective 48
Multi-Source Sparse Interpolated Experts: Task-Conditioned Composition of Multiple Pretrained Priors with Tight Statistical Guarantees Liz Lemma & Future Detective 50 AutoBudget-SMAT: Primal–Dual Meta-Tuning with Learned Per-Module Sparsity/Latency Budgets Liz Lemma & Future Detective 48 Bayes-Routed Sparse Interpolated Experts: PAC-Bayes and Tail-Risk Guarantees for Uncertainty-Aware Meta-Tuning Liz Lemma & Future Detective 38
Certified Multi-Device, Multi-Fidelity Neural Architecture Search for an ε-Pareto Front under Noisy and Drifting Hardware Measurements Liz Lemma & Future Detective 62 Proxy-Only NAS is Information-Theoretically Impossible: Minimal Calibration Conditions for Provable Architecture Selection Liz Lemma & Future Detective 60 Scale-Calibrated Multi-Fidelity NAS for Adapterized Foundation Models: Rank Guarantees and Tight Sample Complexity Liz Lemma & Future Detective 46
Mechanistic Certificates: PAC-Style Robust Causal Guarantees for Interpretability via Component Replacement Liz Lemma & Future Detective 58 Interaction Bases for Tool-Using LLM Agents: Approximation Algorithms, Stability, and Causal Validation Liz Lemma & Future Detective 56 cs_scrape_base-I5-0-2026-01-18_002716 Liz Lemma & Future Detective 55
Inference-Time Scaling Laws for Embodied Agents: Power Laws, Optimality, and Adaptive Deliberation under Latency Constraints Liz Lemma & Future Detective 58 RoboChinchilla: Compute-Optimal Scaling with Simulation and Real-World Data Costs for Robot Foundation Models Liz Lemma & Future Detective 56 Diversity Scaling Laws for Embodied Imitation: Unseen Success Scales with Coverage Radius, Not Episode Count Liz Lemma & Future Detective 50
State-Constrained Offline RL Without Actions: Certified Latent Reachability from Observation-Only Trajectories Liz Lemma & Future Detective 55 Reachability at Scale for State-Constrained Offline RL: Retrieve-then-Certify Graphs with End-to-End Guarantees Liz Lemma & Future Detective 50 Calibrated Reachability for State-Constrained Offline RL in Stochastic MDPs Liz Lemma & Future Detective 48
Certified Shifts-Aware Rewards: Calibration-to-Value Guarantees for Model-Based Offline RL Liz Lemma & Future Detective 60 Likelihood-Free Shifts-Aware Reward Learning for Implicit World Models via f-Divergence Duality Liz Lemma & Future Detective 52 Dual-SAR: Primal–Dual Shift-Aware Rewards for Model-Based Offline RL without Hyperparameter Tuning Liz Lemma & Future Detective 46
Retrieval-Augmented Score Signals for Model-Free Reward-Directed Diffusion RL Liz Lemma & Future Detective 48 Constrained Reward-Directed Diffusion via Primal–Dual Continuous-Time q-Learning with Score Signals Liz Lemma & Future Detective 47 Off-Policy Martingale q-Learning for Reward-Directed Diffusion with Replay and Normalization Constraints Liz Lemma & Future Detective 45
Preprocessing Leakage in Graph Learning Pipelines: A \((T_{\mathrm{pre}}, r_{\mathrm{pre}})\) Audit via Locality Collapse Liz Lemma & Future Detective 62 Rewiring for Computation, Not Just Accuracy: Conductance-Driven Depth Bounds for WL-Style Canonicalization Liz Lemma & Future Detective 48 Budgeted Communication Expressiveness: Unifying Message Passing and Graph Transformers via WL Canonicalization Liz Lemma & Future Detective 45
Meta-Learned Mixture-of-Annotators for Unsupervised Meta-Learning: Tight Bounds via Pairwise Constraint Channels Liz Lemma & Future Detective 58 Sub-Quadratic Task Construction for Unsupervised Meta-Learning: Approximate kNN Graphs and OT Assignments in the Inner Loop Liz Lemma & Future Detective 54 Stability-Guided Meta-Control for Unsupervised Bilevel Meta-Learning Liz Lemma & Future Detective 38
The Proxy Manifold: Recipe-Invariant Downstream Prediction from a Standardized Suite of Losses and Probes Liz Lemma & Future Detective 48 Low-Rank Multi-Source Data-Mixing Laws for LLM Pretraining: Identifiable Scaling with ˜O(rK) Converged Runs Liz Lemma & Future Detective 48 Active-FLP: Correlation-Aware Multi-Fidelity Scaling Laws with Calibrated Prediction Intervals Liz Lemma & Future Detective 48
Mean Aggregation Meets Balanced Optimal Transport: Compact, Separating Metrics for Attributed Graph Neural Networks Liz Lemma & Future Detective 54 Multi-Relational DIDM Mover’s Distance: Compact, Stable, and Universal Theory for Edge-Attributed GNNs Liz Lemma & Future Detective 52 Fast Approximation of DIDM Mover’s Distance via Entropic Multiscale Optimal Transport Liz Lemma & Future Detective 48
On-Manifold Activation Patching: Well-Defined Counterfactuals and Stable Causal Effects in Neural Networks Liz Lemma & Future Detective 63 Bifurcations in Mechanism Space: Change-Point Detection on Interventional Influence Subspaces Predicts Capability Emergence Liz Lemma & Future Detective 48 Hierarchical Residual Sparse Autoencoders: Decomposing SAE Error into a Second Dictionary Liz Lemma & Future Detective 45
Time-Conditioned Denoising Score Matching as Multi-Task Interpolation: Precise Learning Curves and Memorization Thresholds Liz Lemma & Future Detective 62 Compute–Privacy Frontiers in Diffusion Models via Optimal Noise-Reuse Scheduling $m(t)$ Liz Lemma & Future Detective 55 Intrinsic-Dimension Scaling Laws for Memorization in Latent Diffusion via Precise DSM Learning Curves Liz Lemma & Future Detective 46
Iso-Cost Micro-Mutations for Hardware-Aware NAS: Fair Local Search under Latency/Energy Constraints Liz Lemma & Future Detective 58 Ranking-Preserving Coresets for Global NAS via Pairwise-Margin Uniform Convergence Liz Lemma & Future Detective 48 Capacity-Aware Multi-Fidelity Architecture Evaluation: Best-Arm Identification with Proxy Bias and Costs Liz Lemma & Future Detective 45
Loss That Transfers: Calibrated Progress Metrics and Tight Bounds for Cross-Tokenizer Emergence Forecasting Liz Lemma & Future Detective 59 Capability Manifolds for Emergence: Surfaces in Calibrated Pretraining Progress, Post-Training RL, and Test-Time Compute Liz Lemma & Future Detective 58 Critical Subspaces for Emergent Abilities Under Compression: Certified Mixed-Precision Quantization via Margin Sensitivity Liz Lemma & Future Detective 46
Correlated-Time Kernel Expansions for End-to-End Wasserstein Error in Diffusion Models Liz Lemma & Future Detective 62 Compute-Optimal Training and Sampling Schedules for Diffusion Models via Spectral Kernel Error Budgets Liz Lemma & Future Detective 58 No-Sampling Predictors for Diffusion Quality via Spectral Kernels and Training-Time Residuals Liz Lemma & Future Detective 42
Causal Evidence Usage in Retrieval-Augmented LMs via Interchange Interventions Liz Lemma & Future Detective 62 ATR++: Ambiguous Multi-hop Binding as Pointer-Chasing, with Mechanistic Certificates Liz Lemma & Future Detective 62 Mechanistic Generalization Certificates for OOD Retrieval via Causal Signature Invariants Liz Lemma & Future Detective 47
Capacity-Constrained Flow Attention via KL Projections: Sinkhorn/Dykstra-Normalized GNNs with Interpretable Pseudo-Flows Liz Lemma & Future Detective 56 Implicit Flow Attention on Cyclic Graphs: Convergent Conservation-Normalized Message Passing with Stability Bounds Liz Lemma & Future Detective 48 FlowGPS: Conservation-Normalized Local Attention for Scalable Graph Transformers on Flow Networks Liz Lemma & Future Detective 48
Transfer-Aware Regimes for Offline-to-Online RL Under Dynamics Shift Liz Lemma & Future Detective 58 Reset Is a Scalpel: Safe Partial Reset Operators for Offline-to-Online RL with Stability Guarantees Liz Lemma & Future Detective 58 RegimeFlow: Continuous, Uncertainty-Aware Control of Stability–Plasticity in Offline-to-Online RL Liz Lemma & Future Detective 42
Real-Time Sparse Tracing at Million-Token Context: IO-Optimal Compiled STREAM Kernels for Consumer GPUs Liz Lemma & Future Detective 52 PA-STREAM: Position-Adaptive Hierarchical Sparse Tracing for Uniform Long-Context Fidelity Liz Lemma & Future Detective 52 Certified Logit-Faithful Sparse Tracing via Softmax Tail Bounds Liz Lemma & Future Detective 48
Bayes-FORL: Calibrated Bayesian Fusion of Forecast Priors and Offline Generative Beliefs for Non-Stationary Offline RL Liz Lemma & Future Detective 52 Anchors are Necessary: Tight Lower/Upper Bounds for Episodic Sensor Bias in Offline RL Liz Lemma & Future Detective 49 AFORL: Offline RL under Episodic Affine Sensor Drift via Transform-Invariant Belief Updates Liz Lemma & Future Detective 48

Econ

Safe Pacing with Delayed Conversions: High-Probability ROI/CPA Constraints and Liquid Welfare without Convergence Liz Lemma & Future Detective 61 Many-Constraint Autobidding: Ex-Post Feasible Multi-KPI Pacing with Liquid-Welfare Guarantees without Convergence Liz Lemma & Future Detective 58 Liquid Welfare without Convergence under $(\varepsilon,\delta)$-Approximate Core Auctions Liz Lemma & Future Detective 58
Monotone Projections of Log-Linear Pooling: Incentive-Compatible Influence Auctions Aligned with RLHF Liz Lemma & Future Detective 62 Influence Auctions for Agentic LLM Policies: Tool Calls, Citations, and Universal Stable Sampling Liz Lemma & Future Detective 62 Verified Influence Auctions for LLM Markets: Proper-Scoring Audits for Truthful Distribution Reports Liz Lemma & Future Detective 58
Dynamic Incentive Compatibility as Control: Online Monitoring and Stabilization via I-DIC Liz Lemma & Future Detective 58 Inference-Time Feasibility in Neural Auctions via Bregman Projection onto Fairness/Quota Polytopes Liz Lemma & Future Detective 56 Certified Neural Mechanisms: Finite-Sample Guarantees for Approximate IC, IR, and Fairness Liz Lemma & Future Detective 52
Learning Sophistication as Protection: A Robustness Threshold for Dynamic Contracting Against Learning Agents Liz Lemma & Future Detective 62 Auditing Algorithmic Pay: Testable Signatures of Free-Fall Dynamic Contracting against Learning Agents Liz Lemma & Future Detective 62 Contracting with Churn: Dynamic Incentives Against Learning Agents Under Stochastic Stopping Liz Lemma & Future Detective 48
When Does Swap Regret Immunize Privately-Informed Agents? Type-Conditional No-Swap-Regret in Generalized Principal–Agent Problems Liz Lemma & Future Detective 66 Generalized Principal–Agent Design in Nonstationary Environments: A Dynamic Benchmark and Stability Bounds Against Learning Agents Liz Lemma & Future Detective 62 Attention-Constrained Information Design with Learning Agents: Effective Channel Expansion via Actions Liz Lemma & Future Detective 45
No-Free-Anchoring: Platform-Designed Outcome-Weighted Reference Prices and Near-Stationary Optimal Pricing Liz Lemma & Future Detective 62 Measuring Long-Memory Reference Effects from Platform Deal Labels and Price Disclosures: Identification and Counterfactual Platform Design Liz Lemma & Future Detective 58 Memory Matters: Phase Transitions in Dynamic Pricing with Time-Varying Reference Memory $\delta_t \propto (t+1)^{-\alpha}$ Liz Lemma & Future Detective 48
Contract-as-Instrument Breaks Under Selection: Safe Randomization and Online IV Learning with Endogenous Participation Liz Lemma & Future Detective 64 Regulated Linear Contracts in Multitasking: Projection Characterizations and Instrumental Learning Liz Lemma & Future Detective 58 Goodhart-Proof Incentives: Linear Multitask Contracts with Strategic Metric Manipulation and Audits Liz Lemma & Future Detective 56
Intrinsic-VAPE: Contextual Dynamic Pricing on Doubling Metrics with Cross-Learned Demand Liz Lemma & Future Detective 59 Contextual Dynamic Pricing under Demand Drift: Windowed VAPE and Dynamic Regret \(\tilde O(T^{2/3}+T^{2/3}V_T^{1/3})\) Liz Lemma & Future Detective 58 Offline-to-Online VAPE: Conservative Contextual Dynamic Pricing from Biased Logs Liz Lemma & Future Detective 48
Incentive-Compatible Online Mechanism Learning with Endogenous Participation: IR, Selection Bias, and Audits Liz Lemma & Future Detective 62 Continuous-Type Nash Incentive-Compatible Online Mechanism Learning via Random Audits: Stability, $\varepsilon$-NIC, and the Audit–Regret Frontier Liz Lemma & Future Detective 62 Truthful and Adaptive Online Mechanism Learning: Strongly Adaptive Regret under Non-Myopic Manipulation Liz Lemma & Future Detective 58
Robust Nudging in Sequential Contracting: Distributionally Robust Incentive Compatibility for Principal–Agent RL Liz Lemma & Future Detective 60 Contracts for Compute: Incentivizing Verifiable Effort in Tool-Using LLM Agents via Receipt-Contingent Payments Liz Lemma & Future Detective 54 Stabilizing Dynamic Contracts in Infinite Horizon via Entropy-Regularized Subgame-Perfect Equilibrium Liz Lemma & Future Detective 44
The Minimal Coupling Principle for Budget and RoS Pacing: One Shadow Price is Enough (and Sometimes Necessary) Liz Lemma & Future Detective 62 Efficiency Without Convergence: Learning-Robust Liquid Welfare Guarantees for Non-Stationary Autobidding Liz Lemma & Future Detective 59 Platform Competition and Information Design in the Autobidding Era: Why Precision Can Reduce Revenue Liz Lemma & Future Detective 58
Two-Sided Share-Fair Stable Lotteries under One-Sided Ties: Frontiers, Impossibilities, and Nash Compromise Liz Lemma & Future Detective 63 Rolling OSS Schedules: Ex-Post (ϵ-)Internal Stability and Log-Share Fairness under Churn Liz Lemma & Future Detective 42
Why Offline Pricing RL Fails: Non-Identification under Confounding and a Minimal-Randomization Fix with Finite-Sample Bounds Liz Lemma & Future Detective 62 Safe & Stable Dynamic Pricing Under Demand Drift: Offline Pretraining, Change-Point Detection, and Hard-Constraint Learning Guarantees Liz Lemma & Future Detective 60 Constrained Dynamic Pricing with Delayed Returns and Carbon Costs: A Low-Dimensional Pareto Frontier and Implementable Pricing Rules Liz Lemma & Future Detective 45
When Do Synthetic Auction Simulators Support Policy? Validity Bounds for Counterfactual Mechanism Rankings Liz Lemma & Future Detective 64 Transparency vs. Tacit Collusion in Algorithmic Ad Auctions: Optimal Disclosure with Budgets and Learning Bidders Liz Lemma & Future Detective 58 Identification and Sharp Bounds from Top-K Auction Logs in Budgeted Multi-Slot Auctions Liz Lemma & Future Detective 38
Learning Contracts from Telemetry via Implementability Complexity Liz Lemma & Future Detective 64 Hidden Benchmarks and Anti-Gaming: Ambiguous Evaluation with Deployment Commitment Liz Lemma & Future Detective 63 Audit-Augmented Linear Contracts: Rare Verification Restores Approximation Guarantees Liz Lemma & Future Detective 62
Mechanism Design Against Agentic Bidders: Learned Deviation Oracles for Robust Approximate-DSIC Auctions Liz Lemma & Future Detective 62 Robust RegretNet for Combinatorial Auctions: Distributionally-Robust Approximate DSIC Under Wasserstein Shift Liz Lemma & Future Detective 56 Solver-Backed Neural Combinatorial Auctions: Exact Feasibility, Auditable Allocations, and Regret-Stability Bounds Liz Lemma & Future Detective 48
Verified Incentives in Hierarchical Bandits: Action-Contingent Transfers with Random Attestations Liz Lemma & Future Detective 58 Contextual-MAIL: Decentralized Incentivized Learning in Hierarchical Principal–Agent Contextual Bandits Liz Lemma & Future Detective 58 Budget-Feasible Incentivized Learning in Hierarchical Principal–Agent Trees: Shadow Prices, Constrained Efficiency, and Primal–Dual No-Regret Liz Lemma & Future Detective 42
The Price of Privacy in Real-Time Federated Learning: Second-Best Incentives with AoI and Service Latency Signals Liz Lemma & Future Detective 58 Carbon- and Congestion-Priced Federated Learning: Equilibrium Update Frequency and Local Epochs with AoI/Latency Liz Lemma & Future Detective 55 Stakes, Slashing, and Freshness: Adversary-Robust Incentives for Timely Federated Learning in IIoT Liz Lemma & Future Detective 48
Sybil- and Poisoning-Robust Budget-Feasible Mechanisms for Federated Learning with Auditable Compatibility Liz Lemma & Future Detective 57 Dual-Budget Compatibility-Aware Procurement for Federated Learning: Monetary \& Carbon Budgets under Congestion Caps Liz Lemma & Future Detective 52 DriftPay: Budget-Feasible Truthful Procurement for Federated Learning via Model-Contrastive Representation-Drift Scores Liz Lemma & Future Detective 46
Equilibrium-Stable Fairness in Strategic Classification: Selectivity Fails, Inclusiveness Survives Liz Lemma & Future Detective 63 The Transparency Tax in Strategic Classification: Quantifying the Cost of Fully Targetable Decision Rules Liz Lemma & Future Detective 60 Causal Recourse Contracts: Implementing Genuine Improvement and Blocking Proxy Gaming Liz Lemma & Future Detective 58
Learning How Much to Trust: Exogenously-Calibrated Meta-Mechanisms for Truthful Budget-Feasible Procurement Liz Lemma & Future Detective 63
Attention-Budget Auctions for Generative Content: Truthful Matching Under Trust and Compliance Constraints Liz Lemma & Future Detective 64 Blueprint Auctions for LLM Interfaces: Efficient Mechanisms When the Platform Chooses the Generated Narrative Liz Lemma & Future Detective 62 Robust Learning-Aware Position Auctions for LLM-Estimated CTRs via LCB Optimization and Monotone Exploration Liz Lemma & Future Detective 58
Auditable Stochastic Contracts: Anytime-Valid Fairness Certification from Logged Adaptive Contracting Liz Lemma & Future Detective 56 Fairness Without λ: Constrained Linear Contracting and Two-Timescale Convergence in Principal–Agent Markov Games Liz Lemma & Future Detective 52 Safe Fair Contracts in Principal–Agent Markov Games: Minimum Guarantees for Practical Individual Rationality Liz Lemma & Future Detective 42
Receding-Horizon Gaussian-Process Bayesian Optimization for Inventory-Constrained Dynamic Pricing Liz Lemma & Future Detective 56 Tracking Demand Drift in Dynamic Pricing via Sliding-Window Gaussian-Process UCB Liz Lemma & Future Detective 52 Governed GP-UCB for Dynamic Pricing: Regret Guarantees under Fairness and Volatility Constraints Liz Lemma & Future Detective 48
The Merchant and the Librarian Liz Lemma & Future Detective — Search Triggering by Chatbots as Value-of-Information under Misaligned Objectives Liz Lemma & Future Detective — Query Steering in Agentic Search: An Information-Design Model of Monetization Misalignment Liz Lemma & Future Detective — End-to-End vs.~Modular Training for Agentic Search: A Price-of-Anarchy Theory for Chatbot Tool Use Liz Lemma & Future Detective — Audited Search for Agentic Chatbots: Quantitative Bounds on Monetization-Induced Over-Triggering Liz Lemma & Future Detective —

Math

Density Waves for Haar-Unitary Matrices: A Gaussian Free Field Limit and a Zeros-Only Reconstruction Liz Lemma & Future Detective 46 Window Universality for Carrier-Wave Estimators of CUE Characteristic Polynomials Liz Lemma & Future Detective 33
Fraïssé–Nominal Probability and Commutative Monads for Exchangeable Relational Interfaces Liz Lemma & Future Detective 55 Robust Program Semantics for Random Graph Interfaces via Total Variation and Cut Distance Liz Lemma & Future Detective 45 Aldous–Hoover by Program Equations: k-Uniform Hypergraphs, Markov Categories, and Hypergraphons Liz Lemma & Future Detective 45
Mesoscopic Compound-Poisson Approximation for Cycles in Wreath Products Γn ⋊Sn Liz Lemma & Future Detective 48 A Logarithmic Universality Principle for Cycle Counts in Wreath Products Γn ⋊Hn Liz Lemma & Future Detective 48 A Lévy–Khintchine Description of Wreath-Product Cycle Counts Liz Lemma & Future Detective 45
Horizontal Generic Vanishing for Unitary Rank-2 Local Systems under Isomonodromy Liz Lemma & Future Detective 47 Finite \(\mathrm{PMod}_{0,n}\)-Orbits on \(\mathrm{SL}_2\)-Character Varieties for \(n=5,6\) Without Genericity Assumptions Liz Lemma & Future Detective 45 Integral Isomonodromy for Motivic Picard–Fuchs Summands Liz Lemma & Future Detective —
Certified Computation of \(\mathrm{SPP}(7)\) via Addition/Multiplication Table Types Liz Lemma & Future Detective 48 SumProductAtlas 2.0: a proof-producing database of sum–product pairs up to n ≤Nmax with completeness in structural strips Liz Lemma & Future Detective 35 Untitled Liz Lemma & Future Detective 22
A Pointed Completion of the Freiman Category Restoring Finite (Co)limits and Translation Symmetry Liz Lemma & Future Detective 52 Quantitative Stability of Small Doubling Under Pushouts in FR0 2 Liz Lemma & Future Detective 35 Universal Ambient Groups as a Reflection and Monad in the Freiman Category Liz Lemma & Future Detective —
Lorentz Control in Additive Combinatorics: Tail-Free L3 Estimates and Weighted Propagation Liz Lemma & Future Detective 48 Weak L3-control versus full control in additive combinatorics Liz Lemma & Future Detective 38
Many-Frequency Synchronization Below BBP: A λcomp(L) ≍L−1/2 Law via Frequency-Coupled AMP and Growing-L Low-Degree Bounds Liz Lemma & Future Detective 47 Exact Phase Transition at n ≈d2/4 for Centered Ellipsoid Fitting of Gaussian Points Liz Lemma & Future Detective 38 Regularization-to-Exactness for PSD Feasibility under Rank-One Measurements: Polishing Approximate Ellipsoid Fits Liz Lemma & Future Detective —
Tautological Chow Rings and CKgP for \(\mathcal{M}_{5,n}\) in Genus 5 up to \(n=12\) Liz Lemma & Future Detective 57
# Title Folder Score
1 When Does Swap Regret Immunize Privately-Informed Agents? Type-Conditional No-Swap-Regret in Generalized Principal–Agent Problems Generalized Principal-Agent Problem w... 66
2 When Do Synthetic Auction Simulators Support Policy? Validity Bounds for Counterfactual Mechanism Rankings AuctionNet: A Novel Benchmark for Dec... 64
3 Local $k$-Folklore GNNs Beyond Strong Nesting: A $k$-Almost-Strong Nested Ear Decomposition Characterization Beyond Weisfeiler-Lehman: A Quantitat... 64
4 Learning Contracts from Telemetry via Implementability Complexity Algorithmic Contract Theory: A Survey 64
5 Contract-as-Instrument Breaks Under Selection: Safe Randomization and Online IV Learning with Endogenous Participation Linear Contracts in Multitasking: Rob... 64
6 Causal Direct Preference Optimization with Endogenous Prompts Direct Preference Optimization: Your ... 64
7 Attention-Budget Auctions for Generative Content: Truthful Matching Under Trust and Compliance Constraints Position Auctions in AI-Generated Con... 64
8 Two-Sided Share-Fair Stable Lotteries under One-Sided Ties: Frontiers, Impossibilities, and Nash Compromise Stable Matching with Ties: Approximat... 63
9 Preconditioner-Learned Denoising Diffusion Operators: Dimension-Free Acceleration via Learnable Trace-Class Covariances Score-based Diffusion Models in Funct... 63
10 On-Manifold Activation Patching: Well-Defined Counterfactuals and Stable Causal Effects in Neural Networks Open Problems in Mechanistic Interpre... 63
11 Learning How Much to Trust: Exogenously-Calibrated Meta-Mechanisms for Truthful Budget-Feasible Procurement Online Budget-Feasible Mechanism Desi... 63
12 JFT-Exp: Quantitative Joint Feature–Topology Expressivity for GNNs and Graph Transformers The Expressive Power of Graph Neural ... 63
13 Hidden Benchmarks and Anti-Gaming: Ambiguous Evaluation with Deployment Commitment Algorithmic Contract Theory: A Survey 63
14 Equilibrium-Stable Fairness in Strategic Classification: Selectivity Fails, Inclusiveness Survives Incentive-Aware Machine Learning; Rob... 63
15 Constraint-Preserving Denoising Diffusion Operators on Function Spaces Score-based Diffusion Models in Funct... 63
16 Beyond WL for Sparse Graph Transformers: Homomorphism Expressivity Under Attention Constraints Beyond Weisfeiler-Lehman: A Quantitat... 63
17 Why Offline Pricing RL Fails: Non-Identification under Confounding and a Minimal-Randomization Fix with Finite-Sample Bounds Dynamic Retail Pricing via Q-Learning... 62
18 Vector-Lagrangian Safe RLHF: Multi-Category Risk Budgets and Shadow Prices for LLM Safety Governance Constitutional AI: Harmlessness from ... 62
19 Time-Conditioned Denoising Score Matching as Multi-Task Interpolation: Precise Learning Curves and Memorization Thresholds Denoising Score Matching with Random ... 62
20 The Minimal Coupling Principle for Budget and RoS Pacing: One Shadow Price is Enough (and Sometimes Necessary) Auto-Bidding and Auctions in Online A... 62
21 The Independence Premium: Audit Portfolios, Correlated Blind Spots, and Tail-Risk Bounds in AI Oversight Constitutional AI: Harmlessness from ... 62
22 Robust Doubly-Efficient Debate with Correlated, Contaminated Human Judgment Scalable AI Safety via Doubly-Efficie... 62
23 Preprocessing Leakage in Graph Learning Pipelines: A \((T_{\mathrm{pre}}, r_{\mathrm{pre}})\) Audit via Locality Collapse Rethinking GNN Expressive Power from ... 62
24 No-Free-Anchoring: Platform-Designed Outcome-Weighted Reference Prices and Near-Stationary Optimal Pricing Dynamic Pricing and Learning with Lon... 62
25 Monotone Projections of Log-Linear Pooling: Incentive-Compatible Influence Auctions Aligned with RLHF Mechanism Design for Large Language M... 62
26 Mechanism Design Against Agentic Bidders: Learned Deviation Oracles for Robust Approximate-DSIC Auctions Advancing Differentiable Economics: A... 62
27 Learning Sophistication as Protection: A Robustness Threshold for Dynamic Contracting Against Learning Agents Contracting with a Learning Agent 62
28 Influence Auctions for Agentic LLM Policies: Tool Calls, Citations, and Universal Stable Sampling Mechanism Design for Large Language M... 62
29 Incentive-Compatible Online Mechanism Learning with Endogenous Participation: IR, Selection Bias, and Audits Nash Incentive-compatible Online Mech... 62
30 Generalized Principal–Agent Design in Nonstationary Environments: A Dynamic Benchmark and Stability Bounds Against Learning Agents Generalized Principal-Agent Problem w... 62
31 Debate-Triggered Audits for Tool-Using Agents: Closed-Form Oversight Savings, Identification, and 2026-Scale Evidence Scalable AI Safety via Doubly-Efficie... 62
32 Correlated-Time Kernel Expansions for End-to-End Wasserstein Error in Diffusion Models From Score Matching to Diffusion: A F... 62
33 Continuous-Type Nash Incentive-Compatible Online Mechanism Learning via Random Audits: Stability, $\varepsilon$-NIC, and the Audit–Regret Frontier Nash Incentive-compatible Online Mech... 62
34 Certified Multi-Device, Multi-Fidelity Neural Architecture Search for an ε-Pareto Front under Noisy and Drifting Hardware Measurements Evolution and Efficiency in Neural Ar... 62
35 Causal Evidence Usage in Retrieval-Augmented LMs via Interchange Interventions Mechanistic Evaluation of In-Context ... 62
36 Blueprint Auctions for LLM Interfaces: Efficient Mechanisms When the Platform Chooses the Generated Narrative Position Auctions in AI-Generated Con... 62
37 Auditing Algorithmic Pay: Testable Signatures of Free-Fall Dynamic Contracting against Learning Agents Contracting with a Learning Agent 62
38 Audit-Augmented Linear Contracts: Rare Verification Restores Approximation Guarantees Algorithmic Contract Theory: A Survey 62
39 ATR++: Ambiguous Multi-hop Binding as Pointer-Chasing, with Mechanistic Certificates Mechanistic Evaluation of In-Context ... 62
40 Safe Pacing with Delayed Conversions: High-Probability ROI/CPA Constraints and Liquid Welfare without Convergence Autobidders with Budget and ROI Const... 61
41 Debate-as-Compliance: Audit-Triggered Mechanisms for Constant-Cost Oversight of Agentic AI Scalable AI Safety via Doubly-Efficie... 61
42 The Transparency Tax in Strategic Classification: Quantifying the Cost of Fully Targetable Decision Rules Incentive-Aware Machine Learning; Rob... 60
43 Safe & Stable Dynamic Pricing Under Demand Drift: Offline Pretraining, Change-Point Detection, and Hard-Constraint Learning Guarantees Dynamic Retail Pricing via Q-Learning... 60
44 Robust Nudging in Sequential Contracting: Distributionally Robust Incentive Compatibility for Principal–Agent RL Principal-Agent Reinforcement Learnin... 60
45 Proxy-Only NAS is Information-Theoretically Impossible: Minimal Calibration Conditions for Provable Architecture Selection Evolution and Efficiency in Neural Ar... 60
46 No Overlap, No Alignment: Identification Limits for RLHF Reward Models under Endogenous Prompts Goal Misgeneralization: Why Correct S... 60
47 Certified Shifts-Aware Rewards: Calibration-to-Value Guarantees for Model-Based Offline RL Mitigating Distribution Shift in Mode... 60
48 Loss That Transfers: Calibrated Progress Metrics and Tight Bounds for Cross-Tokenizer Emergence Forecasting Emergent Abilities in Large Language ... 59
49 Intrinsic-VAPE: Contextual Dynamic Pricing on Doubling Metrics with Cross-Learned Demand Improved Algorithms for Contextual Dy... 59
50 Efficiency Without Convergence: Learning-Robust Liquid Welfare Guarantees for Non-Stationary Autobidding Auto-Bidding and Auctions in Online A... 59
51 Verified Influence Auctions for LLM Markets: Proper-Scoring Audits for Truthful Distribution Reports Mechanism Design for Large Language M... 58
52 Verified Incentives in Hierarchical Bandits: Action-Contingent Transfers with Random Attestations Learning Contracts in Hierarchical Mu... 58
53 Verifiable Process Certificates as Market Design for AI Safety: A Signaling Model with Alignment-Faking The Alignment Problem from a Deep Lea... 58
54 Truthful and Adaptive Online Mechanism Learning: Strongly Adaptive Regret under Non-Myopic Manipulation Nash Incentive-compatible Online Mech... 58
55 Transparency vs. Tacit Collusion in Algorithmic Ad Auctions: Optimal Disclosure with Budgets and Learning Bidders AuctionNet: A Novel Benchmark for Dec... 58
56 Transfer-Aware Regimes for Offline-to-Online RL Under Dynamics Shift The Three Regimes of Offline-to-Onlin... 58
57 The Price of Privacy in Real-Time Federated Learning: Second-Best Incentives with AoI and Service Latency Signals Meta-Computing Enhanced Federated Lea... 58
58 Robust Learning-Aware Position Auctions for LLM-Estimated CTRs via LCB Optimization and Monotone Exploration Position Auctions in AI-Generated Con... 58
59 Reset Is a Scalpel: Safe Partial Reset Operators for Offline-to-Online RL with Stability Guarantees The Three Regimes of Offline-to-Onlin... 58
60 Regulated Linear Contracts in Multitasking: Projection Characterizations and Instrumental Learning Linear Contracts in Multitasking: Rob... 58
61 Platform Competition and Information Design in the Autobidding Era: Why Precision Can Reduce Revenue Auto-Bidding and Auctions in Online A... 58
62 Overlap as a Design Variable: Optimal Interventions for Robust Reward Modeling and Downstream Alignment Goal Misgeneralization: Why Correct S... 58
63 Optimal Overlap Design for Preference Learning: Quadrant-Filling Interventions and Condition-Number Sample Complexity Direct Preference Optimization: Your ... 58
64 Meta-Learned Mixture-of-Annotators for Unsupervised Meta-Learning: Tight Bounds via Pairwise Constraint Channels Unsupervised Meta-Learning via Dynami... 58
65 Mechanistic Certificates: PAC-Style Robust Causal Guarantees for Interpretability via Component Replacement Mechanistic Interpretability for AI S... 58
66 Measuring Long-Memory Reference Effects from Platform Deal Labels and Price Disclosures: Identification and Counterfactual Platform Design Dynamic Pricing and Learning with Lon... 58
67 Many-Constraint Autobidding: Ex-Post Feasible Multi-KPI Pacing with Liquid-Welfare Guarantees without Convergence Autobidders with Budget and ROI Const... 58
68 Liquid Welfare without Convergence under $(\varepsilon,\delta)$-Approximate Core Auctions Autobidders with Budget and ROI Const... 58
69 Iso-Cost Micro-Mutations for Hardware-Aware NAS: Fair Local Search under Latency/Energy Constraints Efficient Global Neural Architecture ... 58
70 Inference-Time Scaling Laws for Embodied Agents: Power Laws, Optimality, and Adaptive Deliberation under Latency Constraints Neural Scaling Laws in Robotics 58
71 Dynamic Incentive Compatibility as Control: Online Monitoring and Stabilization via I-DIC Deep Learning Meets Mechanism Design:... 58
72 Debate as Anti-Amplification: Competitive Persuasion with Verifiable Evidence and Weak Judges Preference Learning Algorithms Do Not... 58
73 Contextual-MAIL: Decentralized Incentivized Learning in Hierarchical Principal–Agent Contextual Bandits Learning Contracts in Hierarchical Mu... 58
74 Contextual Dynamic Pricing under Demand Drift: Windowed VAPE and Dynamic Regret \(\tilde O(T^{2/3}+T^{2/3}V_T^{1/3})\) Improved Algorithms for Contextual Dy... 58
75 Compute-Optimal Training and Sampling Schedules for Diffusion Models via Spectral Kernel Error Budgets From Score Matching to Diffusion: A F... 58
76 Causal Recourse Contracts: Implementing Genuine Improvement and Blocking Proxy Gaming Incentive-Aware Machine Learning; Rob... 58
77 CarbonBench-FSL: Cost-Aware Few-Shot Evaluation in the Foundation Model Era A Complete Survey on Contemporary Met... 58
78 Capability Manifolds for Emergence: Surfaces in Calibrated Pretraining Progress, Post-Training RL, and Test-Time Compute Emergent Abilities in Large Language ... 58
79 Beyond 1-WL at Scale: Sparse Graph Transformers with Stable Spectral Positional Encoding The Expressive Power of Graph Neural ... 58
80 Audits, Obfuscation, and Alignment-Faking: A Tractable Principal–Agent Model of Frontier AI Compliance The Alignment Problem from a Deep Lea... 58
81 Alignment is a Public Good: Competitive Underinvestment in Overlap for RLHF-Robust Agentic LLMs Goal Misgeneralization: Why Correct S... 58
82 UG-SRDP: Calibrated Uncertainty-Gated Diffusion Policies for Offline RL under State Distribution Shift Diffusion Policies for Out-of-Distrib... 57
83 Tautological Chow Rings and CKgP for \(\mathcal{M}_{5,n}\) in Genus 5 up to \(n=12\) On the Chow Rings of the Moduli Space... 57
84 Sybil- and Poisoning-Robust Budget-Feasible Mechanisms for Federated Learning with Auditable Compatibility CARE: Compatibility-Aware Incentive M... 57
85 Robust RegretNet for Combinatorial Auctions: Distributionally-Robust Approximate DSIC Under Wasserstein Shift Advancing Differentiable Economics: A... 56
86 RoboChinchilla: Compute-Optimal Scaling with Simulation and Real-World Data Costs for Robot Foundation Models Neural Scaling Laws in Robotics 56
87 Receding-Horizon Gaussian-Process Bayesian Optimization for Inventory-Constrained Dynamic Pricing Bayesian Optimization for Dynamic Pri... 56
88 Personalized Alignment with Guarantees: Multi-Objective Constrained DPO under Endogenous Prompts and Fairness Caps Direct Preference Optimization: Your ... 56
89 Interaction Bases for Tool-Using LLM Agents: Approximation Algorithms, Stability, and Causal Validation Mechanistic Interpretability for AI S... 56
90 Inference-Time Feasibility in Neural Auctions via Bregman Projection onto Fairness/Quota Polytopes Deep Learning Meets Mechanism Design:... 56
91 Goodhart-Proof Incentives: Linear Multitask Contracts with Strategic Metric Manipulation and Audits Linear Contracts in Multitasking: Rob... 56
92 Capacity-Constrained Flow Attention via KL Projections: Sinkhorn/Dykstra-Normalized GNNs with Interpretable Pseudo-Flows Flow-Attentional Graph Neural Networks 56
93 Auditable Stochastic Contracts: Anytime-Valid Fairness Certification from Logged Adaptive Contracting Fair Contracts in Principal-Agent Gam... 56
94 cs_scrape_base-I5-0-2026-01-18_002716 Mechanistic Interpretability for AI S... 55
95 State-Constrained Offline RL Without Actions: Certified Latent Reachability from Observation-Only Trajectories State-Constrained Offline Reinforceme... 55
96 Measuring Strategic Misbehavior Under Randomized Oversight: Identification with Concealed Tripwires and Noisy Labels The Alignment Problem from a Deep Lea... 55
97 Fraïssé–Nominal Probability and Commutative Monads for Exchangeable Relational Interfaces Probabilistic Programming Interfaces ... 55
98 Compute–Privacy Frontiers in Diffusion Models via Optimal Noise-Reuse Scheduling $m(t)$ Denoising Score Matching with Random ... 55
99 Carbon- and Congestion-Priced Federated Learning: Equilibrium Update Frequency and Local Epochs with AoI/Latency Meta-Computing Enhanced Federated Lea... 55
100 Sub-Quadratic Task Construction for Unsupervised Meta-Learning: Approximate kNN Graphs and OT Assignments in the Inner Loop Unsupervised Meta-Learning via Dynami... 54
101 Optimal Auditing for Alignment: An Audit Index for Label Allocation under Reference-Conditioned Preference Optimization Preference Learning Algorithms Do Not... 54
102 Mean Aggregation Meets Balanced Optimal Transport: Compact, Separating Metrics for Attributed Graph Neural Networks Generalization, Expressivity, and Uni... 54
103 Contracts for Compute: Incentivizing Verifiable Effort in Tool-Using LLM Agents via Receipt-Contingent Payments Principal-Agent Reinforcement Learnin... 54
104 Tracking Demand Drift in Dynamic Pricing via Sliding-Window Gaussian-Process UCB Bayesian Optimization for Dynamic Pri... 52
105 Synthetic Sparsity Pre-Training for Tail/Cold-Start Robust Graph Prompting under Topology Shift A Survey of Few-Shot Learning on Grap... 52
106 Reconstruct-and-Plan: Trajectory Diffusion with Context Reconstruction for Offline Planning Under Missing Data Diffusion Policies for Out-of-Distrib... 52
107 Real-Time Sparse Tracing at Million-Token Context: IO-Optimal Compiled STREAM Kernels for Consumer GPUs Stream: Scaling up Mechanistic Interp... 52
108 PA-STREAM: Position-Adaptive Hierarchical Sparse Tracing for Uniform Long-Context Fidelity Stream: Scaling up Mechanistic Interp... 52
109 One Backbone, Many Graphs (Without Text): Spectral-Alignment Prompts for Multi-Domain Graph Pre-Training A Survey of Few-Shot Learning on Grap... 52
110 Multi-Relational DIDM Mover’s Distance: Compact, Stable, and Universal Theory for Edge-Attributed GNNs Generalization, Expressivity, and Uni... 52
111 Likelihood-Free Shifts-Aware Reward Learning for Implicit World Models via f-Divergence Duality Mitigating Distribution Shift in Mode... 52
112 Fairness Without λ: Constrained Linear Contracting and Two-Timescale Convergence in Principal–Agent Markov Games Fair Contracts in Principal-Agent Gam... 52
113 Dual-Budget Compatibility-Aware Procurement for Federated Learning: Monetary \& Carbon Budgets under Congestion Caps CARE: Compatibility-Aware Incentive M... 52
114 Certified Neural Mechanisms: Finite-Sample Guarantees for Approximate IC, IR, and Fairness Deep Learning Meets Mechanism Design:... 52
115 Bayes-FORL: Calibrated Bayesian Fusion of Forecast Priors and Offline Generative Beliefs for Non-Stationary Offline RL Forecasting in Offline Reinforcement ... 52
116 A Pointed Completion of the Freiman Category Restoring Finite (Co)limits and Translation Symmetry A categorical approach to additive co... 52
117 Reachability at Scale for State-Constrained Offline RL: Retrieve-then-Certify Graphs with End-to-End Guarantees State-Constrained Offline Reinforceme... 50
118 Multi-Source Sparse Interpolated Experts: Task-Conditioned Composition of Multiple Pretrained Priors with Tight Statistical Guarantees Unleashing the Power of Meta-tuning f... 50
119 Diversity Scaling Laws for Embodied Imitation: Unseen Success Scales with Coverage Radius, Not Episode Count Neural Scaling Laws in Robotics 50
120 Constraint-Aware RoBoT: Feasible-Precision@K Guarantees for Training-Free NAS Under Deployment Budgets Robustifying and Boosting Training-Fr... 49
121 Anchors are Necessary: Tight Lower/Upper Bounds for Episodic Sensor Bias in Offline RL Forecasting in Offline Reinforcement ... 49
122 The Reference-Conditioning Trap: A Population Bound on Mis-Ranking Corrections under Fixed-Reference DPO Preference Learning Algorithms Do Not... 48
123 The Proxy Manifold: Recipe-Invariant Downstream Prediction from a Standardized Suite of Losses and Probes Scaling Laws for Predicting Downstrea... 48
124 Stakes, Slashing, and Freshness: Adversary-Robust Incentives for Timely Federated Learning in IIoT Meta-Computing Enhanced Federated Lea... 48
125 Solver-Backed Neural Combinatorial Auctions: Exact Feasibility, Auditable Allocations, and Regret-Stability Bounds Advancing Differentiable Economics: A... 48
126 Sketch-WL: Near-Linear Randomized Emulation of $k$-WL Statistics via Anchors and Tensor/Count Sketches The Expressive Power of Graph Neural ... 48
127 Rewiring for Computation, Not Just Accuracy: Conductance-Driven Depth Bounds for WL-Style Canonicalization Rethinking GNN Expressive Power from ... 48
128 Retrieval-Augmented Score Signals for Model-Free Reward-Directed Diffusion RL Reward-Directed Score-Based Diffusion... 48
129 Refusal Externalities: Non-Evasive Refusals as an Information Public Good in Dynamic AI Safety Constitutional AI: Harmlessness from ... 48
130 Ranking-Preserving Coresets for Global NAS via Pairwise-Margin Uniform Convergence Efficient Global Neural Architecture ... 48
131 One-Round Federated Few-Shot Personalization via Prompt–Adapter Meta-Learning with Communication Lower Bounds A Complete Survey on Contemporary Met... 48
132 Offline-to-Online VAPE: Conservative Contextual Dynamic Pricing from Biased Logs Improved Algorithms for Contextual Dy... 48
133 Mesoscopic Compound-Poisson Approximation for Cycles in Wreath Products Γn ⋊Sn Analysis of Clustering and Degree Ind... 48
134 Memory Matters: Phase Transitions in Dynamic Pricing with Time-Varying Reference Memory $\delta_t \propto (t+1)^{-\alpha}$ Dynamic Pricing and Learning with Lon... 48
135 Low-Rank Multi-Source Data-Mixing Laws for LLM Pretraining: Identifiable Scaling with ˜O(rK) Converged Runs Scaling Laws for Predicting Downstrea... 48
136 Lorentz Control in Additive Combinatorics: Tail-Free L3 Estimates and Weighted Propagation Control and its applications in addit... 48
137 Local-Access Graph Foundation Models: Sublinear Few-Shot Adaptation with Prompt Tokens A Survey of Few-Shot Learning on Grap... 48
138 Implicit Flow Attention on Cyclic Graphs: Convergent Conservation-Normalized Message Passing with Stability Bounds Flow-Attentional Graph Neural Networks 48
139 Governed GP-UCB for Dynamic Pricing: Regret Guarantees under Fairness and Volatility Constraints Bayesian Optimization for Dynamic Pri... 48
140 FlowGPS: Conservation-Normalized Local Attention for Scalable Graph Transformers on Flow Networks Flow-Attentional Graph Neural Networks 48
141 Fast Approximation of DIDM Mover’s Distance via Entropic Multiscale Optimal Transport Generalization, Expressivity, and Uni... 48
142 Distribution-Free Certificates for Top-K Exploitation in Training-Free NAS Robustifying and Boosting Training-Fr... 48
143 Contracting with Churn: Dynamic Incentives Against Learning Agents Under Stochastic Stopping Contracting with a Learning Agent 48
144 Certified Logit-Faithful Sparse Tracing via Softmax Tail Bounds Stream: Scaling up Mechanistic Interp... 48
145 Certified Computation of \(\mathrm{SPP}(7)\) via Addition/Multiplication Table Types Visualizing the Sum-Product Conjecture 48
146 Calibrated Reachability for State-Constrained Offline RL in Stochastic MDPs State-Constrained Offline Reinforceme... 48
147 Bifurcations in Mechanism Space: Change-Point Detection on Interventional Influence Subspaces Predicts Capability Emergence Open Problems in Mechanistic Interpre... 48
148 AutoBudget-SMAT: Primal–Dual Meta-Tuning with Learned Per-Module Sparsity/Latency Budgets Unleashing the Power of Meta-tuning f... 48
149 ApproxRoBoT: Training-Free NAS in Implicit Spaces via Quantile Candidate Generators Robustifying and Boosting Training-Fr... 48
150 Active-FLP: Correlation-Aware Multi-Fidelity Scaling Laws with Calibrated Prediction Intervals Scaling Laws for Predicting Downstrea... 48
151 AFORL: Offline RL under Episodic Affine Sensor Drift via Transform-Invariant Belief Updates Forecasting in Offline Reinforcement ... 48
152 A Logarithmic Universality Principle for Cycle Counts in Wreath Products Γn ⋊Hn Analysis of Clustering and Degree Ind... 48
153 Mechanistic Generalization Certificates for OOD Retrieval via Causal Signature Invariants Mechanistic Evaluation of In-Context ... 47
154 Many-Frequency Synchronization Below BBP: A λcomp(L) ≍L−1/2 Law via Frequency-Coupled AMP and Growing-L Low-Degree Bounds Randomstrasse101: Open Problems of 2024 47
155 Horizontal Generic Vanishing for Unitary Rank-2 Local Systems under Isomonodromy Motives, mapping class groups, and mo... 47
156 Constrained Reward-Directed Diffusion via Primal–Dual Continuous-Time q-Learning with Score Signals Reward-Directed Score-Based Diffusion... 47
157 Scale-Calibrated Multi-Fidelity NAS for Adapterized Foundation Models: Rank Guarantees and Tight Sample Complexity Evolution and Efficiency in Neural Ar... 46
158 Intrinsic-Dimension Scaling Laws for Memorization in Latent Diffusion via Precise DSM Learning Curves Denoising Score Matching with Random ... 46
159 Dual-SAR: Primal–Dual Shift-Aware Rewards for Model-Based Offline RL without Hyperparameter Tuning Mitigating Distribution Shift in Mode... 46
160 DriftPay: Budget-Feasible Truthful Procurement for Federated Learning via Model-Contrastive Representation-Drift Scores CARE: Compatibility-Aware Incentive M... 46
161 Density Waves for Haar-Unitary Matrices: A Gaussian Free Field Limit and a Zeros-Only Reconstruction Currently there are no reasons to dou... 46
162 Critical Subspaces for Emergent Abilities Under Compression: Certified Mixed-Precision Quantization via Margin Sensitivity Emergent Abilities in Large Language ... 46
163 Robust Program Semantics for Random Graph Interfaces via Total Variation and Cut Distance Probabilistic Programming Interfaces ... 45
164 Real-Time SRDP via Few-Step Distillation: Tight Performance Bounds from Teacher-Student Action Divergence Diffusion Policies for Out-of-Distrib... 45
165 Pareto-Optimal Few-Shot Adaptation in 2026: Cost-Regularized In-Context Learning + Retrieval + Low-Rank PEFT under Cross-Domain Shift A Complete Survey on Contemporary Met... 45
166 Off-Policy Martingale q-Learning for Reward-Directed Diffusion with Replay and Normalization Constraints Reward-Directed Score-Based Diffusion... 45
167 Hierarchical Residual Sparse Autoencoders: Decomposing SAE Error into a Second Dictionary Open Problems in Mechanistic Interpre... 45
168 Finite \(\mathrm{PMod}_{0,n}\)-Orbits on \(\mathrm{SL}_2\)-Character Varieties for \(n=5,6\) Without Genericity Assumptions Motives, mapping class groups, and mo... 45
169 Constrained Dynamic Pricing with Delayed Returns and Carbon Costs: A Low-Dimensional Pareto Frontier and Implementable Pricing Rules Dynamic Retail Pricing via Q-Learning... 45
170 Capacity-Aware Multi-Fidelity Architecture Evaluation: Best-Arm Identification with Proxy Bias and Costs Efficient Global Neural Architecture ... 45
171 Budgeted Communication Expressiveness: Unifying Message Passing and Graph Transformers via WL Canonicalization Rethinking GNN Expressive Power from ... 45
172 Attention-Constrained Information Design with Learning Agents: Effective Channel Expansion via Actions Generalized Principal-Agent Problem w... 45
173 Aldous–Hoover by Program Equations: k-Uniform Hypergraphs, Markov Categories, and Hypergraphons Probabilistic Programming Interfaces ... 45
174 A Lévy–Khintchine Description of Wreath-Product Cycle Counts Analysis of Clustering and Degree Ind... 45
175 Stabilizing Dynamic Contracts in Infinite Horizon via Entropy-Regularized Subgame-Perfect Equilibrium Principal-Agent Reinforcement Learnin... 44
176 Safe Fair Contracts in Principal–Agent Markov Games: Minimum Guarantees for Practical Individual Rationality Fair Contracts in Principal-Agent Gam... 42
177 Rolling OSS Schedules: Ex-Post (ϵ-)Internal Stability and Log-Share Fairness under Churn Stable Matching with Ties: Approximat... 42
178 Resolution-Invariant Error Bounds for Denoising Diffusion Operators via Function-Space Langevin Stability Score-based Diffusion Models in Funct... 42
179 RegimeFlow: Continuous, Uncertainty-Aware Control of Stability–Plasticity in Offline-to-Online RL The Three Regimes of Offline-to-Onlin... 42
180 No-Sampling Predictors for Diffusion Quality via Spectral Kernels and Training-Time Residuals From Score Matching to Diffusion: A F... 42
181 Budget-Feasible Incentivized Learning in Hierarchical Principal–Agent Trees: Shadow Prices, Constrained Efficiency, and Primal–Dual No-Regret Learning Contracts in Hierarchical Mu... 42
182 Weak L3-control versus full control in additive combinatorics Control and its applications in addit... 38
183 Stability-Guided Meta-Control for Unsupervised Bilevel Meta-Learning Unsupervised Meta-Learning via Dynami... 38
184 Identification and Sharp Bounds from Top-K Auction Logs in Budgeted Multi-Slot Auctions AuctionNet: A Novel Benchmark for Dec... 38
185 Exact Phase Transition at n ≈d2/4 for Centered Ellipsoid Fitting of Gaussian Points Randomstrasse101: Open Problems of 2024 38
186 Bayes-Routed Sparse Interpolated Experts: PAC-Bayes and Tail-Risk Guarantees for Uncertainty-Aware Meta-Tuning Unleashing the Power of Meta-tuning f... 38
187 SumProductAtlas 2.0: a proof-producing database of sum–product pairs up to n ≤Nmax with completeness in structural strips Visualizing the Sum-Product Conjecture 35
188 Quantitative Stability of Small Doubling Under Pushouts in FR0 2 A categorical approach to additive co... 35
189 Window Universality for Carrier-Wave Estimators of CUE Characteristic Polynomials Currently there are no reasons to dou... 33
190 Untitled Visualizing the Sum-Product Conjecture 22
191 When Does Homomorphism Expressivity Exist? A Meta-Theorem for Stabilizing Refinement GNNs Beyond Weisfeiler-Lehman: A Quantitat... —
192 Universal Ambient Groups as a Reflection and Monad in the Freiman Category A categorical approach to additive co... —
193 The Merchant and the Librarian alignment/research —
194 The Merchant and the Librarian econ/research —
195 Search Triggering by Chatbots as Value-of-Information under Misaligned Objectives alignment/research —
196 Search Triggering by Chatbots as Value-of-Information under Misaligned Objectives econ/research —
197 Regularization-to-Exactness for PSD Feasibility under Rank-One Measurements: Polishing Approximate Ellipsoid Fits Randomstrasse101: Open Problems of 2024 —
198 Query Steering in Agentic Search: An Information-Design Model of Monetization Misalignment alignment/research —
199 Query Steering in Agentic Search: An Information-Design Model of Monetization Misalignment econ/research —
200 Integral Isomonodromy for Motivic Picard–Fuchs Summands Motives, mapping class groups, and mo... —
201 End-to-End vs.~Modular Training for Agentic Search: A Price-of-Anarchy Theory for Chatbot Tool Use Sponsored Questions and How to Auctio... —
202 End-to-End vs.~Modular Training for Agentic Search: A Price-of-Anarchy Theory for Chatbot Tool Use Sponsored Questions and How to Auctio... —
203 Audited Search for Agentic Chatbots: Quantitative Bounds on Monetization-Induced Over-Triggering alignment/research —
204 Audited Search for Agentic Chatbots: Quantitative Bounds on Monetization-Induced Over-Triggering econ/research —